The word proteome was coined in 1995 to refer to the total protein complement of a genome. The human genome encodes roughly 100,000 genes, corresponding to a similar number of proteins. Not all genes are expressed in all tissues; roughly 10,000 proteins are found in any particular cell. The fraction of the proteome that is expressed by an organism varies between tissues and in response to the environment. Conventional proteome analysis is preformed by two-dimensional gel electrophoresis and requires the protein from roughly a million cells. We have improved the sensitivity of proteome analysis by six orders of magnitude. In preliminary work, we have demonstrated that we can perform a simple analysis of the proteome in a single human cancer cell. Our Preliminary work will be expanded in this R33 proposal. We will automate the manipulations of single cells, we will multiplex the instrument so that 96 cells can be analyzed simultaneously, we will expand our proteome analysis to two-dimensional electrophoresis, and we will evaluate the technology by monitoring the evolution of protein expression in mouse skin tumors. Single cell proteome analysis offers several important advantages. In particular, we can monitor the distribution in the expression of protein markers that are correlated with cancer stage. Like ploidy measurements, the distribution of protein expression may have valuable prognostic value. Sub-populations of metastatic or therapy-resistant cells may be identified at an early stage to guide treatment.